Applicability of the Fridman–Macheret α-Model to Heterogeneous Processes in the Case of Dissociative Adsorption of N₂ on the Ru Surface

Abstract

The efficacy of vibrational excitation for reaction promotion can be well-described by the Fridman–Macheret (F–M) α-model, which was initially developed for processes in the gas phase. However, it is widely used for analysis of the efficacy of vibrational excitation of heterogeneous processes, such as NH3 synthesis, reduction of CO2, and activation of CH4. For plasma-assisted heterogeneous processes, the contribution of species with high vibrational excitation into reaction is emphasized. The aim of this paper is to check the applicability of the F–M model for the case of activation of the N2 molecule on the Ru(113) and Ru(0001) surfaces using molecular dynamics with machine learning potential (MLP). The results show that the F–M model appears to be a reasonable approximation for estimate of vibrational efficacy of highly excited states (v ≥ 7) which are important under nonequilibrium plasma conditions.

Review: Machine learning for advancing low-temperature plasma modeling and simulation

“…Multiple shared MLPs were used to combine the output with information on the electric field, adsorbate energy, and surface energy to predict all quantities of interest (i.e., adsorption energies, dipole moments, and polarizabilities of reactive intermediates). Kedalo et al 103 performed MD simulations to show that the activation of highly excited N 2 molecules on Ru surfaces during NH 3 synthesis can be approximated by the Fridman-Macheret α model. 104 The interatomic interactions were described by an ML potential, utilizing smooth overlap of the atomic positions (SOAP) descriptors, 105 to enable ab initio accuracy but with increased computational efficiency.…”